Re: This should not happen. Contact <address@hidden>. at parallel line 3384.

[Ole Tange]

On Fri, Apr 1, 2011 at 2:33 PM, Sebastian Eiser
<sebastian.eiser@gmail.com> wrote:
> 2011/3/22 Ole Tange <tange@gnu.org>:
>
>>>> * Can you make a set of dummy files that fails too? Maybe a copy of
>>>> the original files but with no content: cp -rs /the/dir my_dir; find
>>>> my_dir -type l | parallel rm {} '&&' touch {}
>
> I cannot make the randomsleep programme fail for the set of dummy
> files, but the original command also fails here. I think it is related
> to the quotes used:

Can you make it fail if you use randomsleep with the arguments that
the original has?

What we are trying to find is an example that you can give to me and
will give the same error on my machine.

If this command fails:

> find my_dir/ -type f -name *.he5 | parallel --eta
> /home/seiser/src/gtproduct-omi-old netcdf -m \'O3.COLUMN\' -fiv
> \'O3.COLUMN\' -f \'"CLOUD.COVER<0.fe \'FLAG.QUALITY.PROCESSING\' -o
> /dev/shm/asdf {}

Then what about this (where randomsleep is just prepended):

> find my_dir/ -type f -name *.he5 | parallel --eta
> randomsleep /home/seiser/src/gtproduct-omi-old netcdf -m \'O3.COLUMN\' -fiv
> \'O3.COLUMN\' -f \'"CLOUD.COVER<0.fe \'FLAG.QUALITY.PROCESSING\' -o
> /dev/shm/asdf {}

> However, issue 3 does not occur if I remove the quotes of the command,
> find my_dir/ -type f -name *.he5 | parallel --eta
> /home/seiser/src/gtproduct-omi-old netcdf -o /dev/shm/asdf {}
>
> where only issue 1 and 2 ("Use of uninitialized value in hash element
> at /home/seiser/alt/usr/bin/parallel line 3284, <STDIN> line 834." and
> the "{} corruption" appear.
>
> hope this helps,

It does not help a lot: I _need_ an example that fails on my system. I
do not have  /home/seiser/src/gtproduct-omi-old so as long as you can
only reproduce the error when using software I have no access to, this
bug will not be fixed.

So please spend some time on making an example that fails without
depending on special software.

/Ole